Development of Immersion Grating for Mid-Infrared High Dispersion Spectrograph for the 8.2m Subaru Telescope Noboru Ebizuka*a, Shinya Morita a, Tomoyuki Shimizu a, Yutaka Yamagata a , Hitoshi Omori a, Moriaki Wakaki b, Hideomi Kobayashi c, Hitoshi Tokoro c, Yasuhiro Hirahara c a RIKEN (The Institute of Physical and Chemical Research); b Department of Electro-Photo-Optics, Tokai University; c Department of Earth and Planetary Sciences, Nagoya University ABSTRACT The mid-infrared high dispersion spectrograph (IRHS; tentative name) with a resolving power of 200,000 at 10 µm is a candidate of the second-generation instrument for the 8.2m Subaru Telescope. A germanium immersion grating will be employed as a dispersing element for this instrument. Germanium immersion gratings for the prototype IRHS were successfully fabricated by using a nano precision 3D profile grinding/turning machine and ELID grinding method on diamond machining. As a result, the fabricated gratings observed to have grooves with ideal saw-tooth shape, smooth surface and acceptable wave front error of a diffraction beam at 10µm. In the present paper, we characterized the performance of the developed immersion gratings. Key word: Solid grating, GaAs, Laser ablation
1. INTRODUCTION The mid-InfraRed High dispersion Spectrograph (IRHS; tentative name) is under planning design study as a secondgeneration instrument for the 8.2m Subaru telescope [1, 2] on Mauna Kea, Hawaii (Fig.1). IRHS is aiming resolving power of λ/∆λ= 200,000 at 10µm (Fig.2), in order to observe vibrational transitions of molecule in circum-stellar and dark clouds for instance. Such a high dispersion spectrograph requires a dispersing element or moving mirror with minimum 2m in an optical path difference. A Fourier transform spectrometer (FTS) is usually adopted as a high dispersion spectrometer for laboratory use from middle to far infrared because of the small size and low cost of the instrument compared with the dispersing type. Although FTS has such disadvantages as background noise and atmospheric turbulence, only several hundreds or tens astronomical objects in the total sky can be observed around 10µm. On the other hand, a mid-infrared high dispersion spectrograph with a conventional reflection grating (Fig.3a) as the dispersing element requires the installation of large optics. Since a collimated beam of 40cm in diameter should be used for the conventional grating spectrograph with a resolving power of 200,000 at 10µm, the total volume of the instrument covered with a below 50K cooled vacuum chamber is nearly 100m3. It is huge enough even for the Nathmith focus of the Subaru telescope.
Fig. 1 8.2m Subaru Telescope.
Fig. 2 Resolving power versus wavelength of Instruments for Subaru Telescope
____________________________________________________________________
* RIKEN (The Institute of Physical and Chemical Research), Wako, Saitama 351-0198 Japan;
[email protected]; http://optik2.mtk.nao.ac.jp/~ebi/ebi-e.html, http://atlas.riken.go.jp/~ebizuka/ebi.html (Japanese)
L
L
L ns
ΔL=2L
(a) Reflaction grating
ΔL=(ns-1)L ΔL=2nsL
(b) Grism
ns
(c) Immersion grating
Fig. 3 Schematic representation of gratings
A grism (Fig. 3b) and immersion grating (Fig.3c) which are directly grooved onto high index material have advantages of reduction of size and weight for a spectroscopic instrument. The grism is a direct vision grating that is combined with a prism and transmission grating. A vertex angle of the prism is adjusted to redirect the diffracted beam straight along the optical axis at a specific wavelength. Major benefit of using a grism, instead of a reflection grating (Fig.3a), in an astronomical instrument is the possibility for allowing flexible switching between imaging and spectroscopic mode by simply inserting or removing the grism in the light beam. The immersion grating is a reflection grating of which optical pass is filled up a dielectric medium. A germanium immersion grating has four times angular dispersion compared with conventional reflection gratings of the same size, since refractive index is over 4.0 in the infrared wavelengths from 1.6 to 20µm and the angular dispersion is proportional to the refractive index of the medium. This means the germanium immersion grating could reduce down to 1/16 as the total volume of a spectrograph compared with a conventional one of a reflection grating. A germanium immersion grating will be used as the dispersing element for IRHS (Fig.4).
2. Fabrication of Immersion Grating 2.1. Germanium single crystal To realize a spectrograph with a resolving power of 200,000 at 10µm, the minimum length of the germanium along the optical axis is 250mm. Figure 5 shows a design of an immersion grating. A germanium single crystal of 400mm in diameter and 120mm in thickness had grown by TDY Co. of Japan. We obtained germanium blocks of 32 x 32 x 80 mm manufactured by the same method as the crystal growth of the germanium of 400mm in diameter. The optical quality of the block was investigated using an InSb infrared camera applying for the projection and infrared interferometer. A small amount of absorbing striation is seen in the projection image (Fig. 6), but no wave front error is observed with the interferometer. The quality of germanium block is thus found to be suitable for use as an immersion grating.
Fig. 4 Schematic representation of mid-infrared high dispersion spectrograph
591
.5
214.4
120
Resolving Power R=200,000 could be accomplished by use of the Immersion grating. 441th Order at 10µm 551.2
[µm]
270
105
120
68.75゜
n=4.00 Fig. 5 Germanium immersion grating for IRHS
Fig. 6 Infrared projection image of single crystal germanium block.
2.2. Trial fabrications for solid gratings In order to fabricate an immersion grating and grism with deep grooves directly onto a high index material that are called solid gratings, numerous studies have been carried out by many researchers and engineers to test various micromachining methods. These include precision ruling-engine, anisotropic chemical etching and so on [3-7]. To our knowledge, only a KRS-5 grism used for near to middle infrared can be fabricated by the use of a precise ruling-engine [6]. Since KRS-5 is a mixed crystal of TlBr and TlI, homogeneous and large blocks are difficult to grow. We have also been carrying out tests of various methods, diamond grinding with resinous bonded tool, ion etching and laser ablation (Left of Fig. 7) [8-13], for example. However solid gratings with deep grooves are difficult to fabricate with these methods (Right of Fig. 7). Finally, we had successfully fabricated ideal grooves onto germanium substrate by using a nano precision 3D profile grinding/turning machine (Fig. 9) and electrolytic in-process dressing (ELID) [14] method for diamond grinding (Fig. 10) [15-18]. 2.3. Fabrication of immersion grating by means of nano precision machine with ELID grinding method The prototype immersion gratings had been designed for a spectrograph with a resolving power of 50,000 at 10µm or 250,000 at 2.0µm. The sizes of the prototype immersion gratings are 30 x 30 mm, 72mm in length and vertex angle of 68.75 degree. Groove spacing are 100µm for the first, 250µm for the second and third, 600µm for the forth
Fig. 7 Schematic representation of grooves formation by means of laser ablation (left) and fabricated grooves on germanium substrate (right).
Fig. 8 Nano precision 3D profile grinding/turning machine
and fifth fabrications. The grooves are slightly tilted to the incident aperture of the gratings to avoid influence of reflection of the incident aperture. The substrate material of the immersion grating was germanium except the third fabrication. At the third fabrication, GaAs single crystal was used for substrate of the immersion grating. The third immersion grating was planed to use for a near infrared imager of InSb that we had already developed, and GaAs was chosen for a substrate material because GaAs is transparent form 0.9µm and refractive index is about 3.4 at 1.0µm. We have used two kinds of the nano precision 3D profile grinding/turning machines for the fabrications of the immersion gratings. The XYZ axes of one machine, which had been used from the first to fourth fabrications, are supported by cross roller bearing and are able to archive 1.0nm in control precision and measurement resolution (Fig. 8). The XYZ axes of the other machine, which had been used at the fifth fabrication, are supported by hydrostatic bearing and are able to archive 0.7 nm in control precision and measurement resolution. The control precision was set 10nm for all of the fabrications. Cast iron bonded diamond-grinding cups of #4,000 and #20,000 were used for the starting and finishing grinding respectively for every fabrications except the second fabrication. At the second fabrication, the diamond-grinding cup of #4,000 was used for the starting grinding and a sintered diamond wheel of #12,000 was used for finishing grinding. The diamond-grinding cups and the sintered diamond wheel were roughly shaped by an electrical spark tooling method with a copper electrode, and sharpened and dressed by a single diamond tooler with ELID grinding method for the first and second fabrications. From the third to fifth fabrications, a rotary tooler of a resin-bonded diamond-grinding wheel of #325 was used for tooling and dressing instead of the copper electrode and single diamond tooler (Fig. 9). The ground depth and speed were determined by test fabrications performed onto a germanium wafer. To evaluate an error of grooves spacing, an on-machine checker of a diffraction image which consists semiconductor laser of 685nm in wavelength, a single mode optical fiber, two camera lens for collimator and imaging, and a CCD camera are used at the finishing grinding at the fourth and fifth fabrication (Fig.10). Figure 11 shows the finished fourth grating.
Fig. 9 Fabrication of prototype germanium immersion grating using nano precision 3D profile grinding/turning machine.
Fig.10 On-machine checker for diffraction image of grating.
3. Performance Evaluations of Immersion Gratings 3.1. Groove shapes Groove shapes of the first and second gratings were not sufficient for practical use. A little amount of beams could come back to the incident surface for the first grating because its grooves are round shape due to the abrasion of the diamond cup. Grooves of the second grating had a lot of fatal chippings. Groove shapes of the third, fourth and fifth gratings are almost ideal saw-tooth shape. Figure 12 shows the cross section of the grooves of the first and fourth grating. 3.2. Wave front image. Figure 13 shows wave front image of the third, fourth and fifth gratings taken by an interferometer of 633nm laser. Because of the wave front error of the third grating is too large, a lot of vacant data points are seen in the interferogram. The wave front error of the fourth and fifth grating are acceptable for a prototype immersion grating at 10 micron if the maximum value of wave front error inside germanium is set up quarter wave in rms, that is, 312.5nm in the air. However, week periodic undulation caused by cross rollers could be seen in the interferogram of the third grating. Such a periodic undulation of wave front error cause false spectrum, namely, Lyman ghost. On the other hand, the wave front of the fifth grating is very smooth but an undulation with large scale is seen. The undulation is supposed to be mainly result of the thermal expansion of the fabricated machine. Such a large-scale undulation becomes the origin of width expansion and split of a diffraction image, namely, parasitic spectrum or Roland ghost.
Fig.11 Prototype germanium immersion grating (30 x 30 x 72 [mm], α = 68.75 [deg.], 1.67 [grooves/mm])
Fig. 12 Grooves shape of the immersion gratings of 100 (above) and 600 (below) µm groove spacing.
3.3. Far field image of a diffraction beam Figure 14 shows far field images of a diffraction beam of the third, fourth and fifth gratings at 10.6µm respectively. A beam of the CO2 laser as a light source at 10.6µm was transformed to a Gaussian beam by means of a spatial filter. Figure 15 shows the cross sections of the far field images of an incident aperture and the diffraction beams of the gratings mentioned above. The solid thin line is a far field image of the reflected beam of the incident aperture which full width half maximum (FWHM) is 90 % thinner than the ideal far field image of grooves because effective area of grooves is 90 % of the incident aperture. The solid, tiny and large doted lines are the far field images of diffracted beams of immersed grooves around 156th order of the third, 453th order of fourth and fifth gratings respectively. The FWHM of the ideal image for the measurement system is 220µm. The FWHM of the images of the third, fourth and fifth gratings are 280, 259 and 302µm respectively. The far field image of the third grating has side lobes with large amplitude caused by fatal wave front error. The FWHM of far field images of the fourth grating expands 18% which is compared with the ideal image, and it is seen a Roland ghost which amplitude is about 17% of the main lobe. The FWHM of the fifth grating expands 37%, and it is seen a Roland ghost which amplitude is about 14% of the main lobe. The expansion and Roland ghost implies a wave front error with large scale of a grating, that is, a deviation of the groove interval.
3rd grating, PV: 961nm, rms: 154nm
3rd grating
4th grating, PV: 583nm, rms: 87nm
4th grating
5th grating, PV: 577nm, rms: 107nm Fig. 13 Wave front images of gratings at 633nm
5th grating Fig.14 Far field images of immersion gratings at 10.6µm
Fig. 15 Cross section of far field images of immersion gratings.
3.4. Surface roughness and scattering Surface roughness of a groove facet for the fourth and fifth gratings was measured by using a surface profiler of contact type. The surface roughness of a groove facet of the fourth and fifth gratings were 11.5 and 3.0 nm in rms respectively. The estimated scatter of the fourth and fifth gratings are 0.3% and 0.002% at 10µm respectively, and 5.3% and 0.36% at 2.5µm respectively. 3.5. Diffraction efficiency To investigate the diffraction efficiency, transmittance of the gratings from 2 to 16µm was measured by using a conventional spectrophotometer. However we had obtained the doubtful values because grooves are tilted about 0.8 degree to the incident aperture, and an incident ray is bent about 3.2 degrees at a groove. We could say that the diffraction efficiency of the third, fourth and fifth gratings are at least 30%. We are preparing a stable light souse and detector at 10µm for measurement of diffraction efficiency.
4. Conclusion and Future Works Although further improvement for the wave front profile should be done, the results of the trial fabrication suggest that a geranium immersion grating used at around 10 µm can be realized by means of the fabrication method, which combine a nano precision 3D profile grinding/turning machine with ELID grinding method. The prototype IRHS is going to be finished by 2002, and the fourth or fifth grating would to be used for the dispersion element of the instrument. The germanium immersion grating shown in figure 3 will be fabricated by 2003. We expect that the construction of the regularIRHS will be completed in 2004.
Fig. 16 Visible (left) and infrared (right) photographs of germanium prism. (120 x 120 x 270 [mm]) Heated solder could see through a germanium of 120 mm in thickness by means of infrared camera.
ACKNOWLEDGMENTS The authors acknowledge professor Dr. Sato of Nagoya University and Dr. Kataza of the Institute of Space and Astronautical Science of Japan for critical comments and continuous encouragement to support this research. We wish to thank colleagues of RIKEN (the Institute of Physical and Chemical Research) for advice on fabricating gratings and their help in using the diamond grinding/turning facility. The work was funded by a grant-in-aid of R&D for the Subaru telescope of the National Astronomical Observatory of Japan.
REFERENCES 1. N. Kaifu et al., "The First Light of the Subaru Telescope: A New Infrared Image of the Orion Nebula", Publ. Astron. Soc. Japan, 52, 1-8 and Plate 1-5, 2000 2. N. Kaifu, "Subaru telescope", Proc. SPIE, 3352, 14-22, 1998. 3. H. Dekker, "An Immersion Grating for an Astronomical Spectrograph", Instrumentation for Ground-Based Optical Astronomy, Present and Future, ed, L.B. Robinson, IGBO.conf. 183, 183-188, 1988 4. G. Wiedemann, ESO Conf. Proc. 40, 305-308, 1992, G. Wiedemann and D.E. Jennings, "Immersion grating for infrared astronomy", Appl. Opt., 32, 1176-1178, 1993 5. H.U. Kaufl and B. Delabre, "Improved Design and Prototyping for a 10/20 mm Camera/Spectro-meter for ESO's VLT", Proc. SPIE, 2198, 1036-1046, 1994 6. M. Shure, D.W. Toomey, J.T. Rayner, P.M. Onaka and A.T. Denault, "NSFCAM: a new infra-red array camera for the NASA Infrared Telescope Facility", Proc. SPIE, 2198, 614-622, 1994 7. L. Weitzel , A. Krabbe, H. Kroker, N. Thatte, L.E. Tacconi-Garman, M. Cameron and R. Genzel, "3D: The next generation near-infrared imaging spectrometer", Astron. Astrophys. Suppl. Ser. 119, 531-546, 1996 8. N. Ebizuka, M. Iye and T. Sasaki, "Optically Anisotropic Crystalline Grisms for Astronomical Spectrographs", Appl.Opt. 37, 1236-1242, 1998. 9 N. Ebizuka, M. Iye, T. Sasaki and M. Wakaki, "Development of High dispersion grisms and immersion gratings for spectrographs of Subaru Telescope", Proc. SPIE, 3355, 409-416, 1998 10. N. Ebizuka, H. Kobayashi, Y. Hirahara, M. Wakaki, K. Kawaguchi, T. Sasaki and M. Iye; "Development of grisms and immersion gratings for the spectrographs of the Subaru Telescope, "Imaging the Universe in Three Dimensions, eds. W. van Breugel and J. Bland-Hawthorn, A.S.P. Conf. Ser. 195, 564-567, 1999 11. Y. Aoyagi and S. Nanba, "Blazed ion-etched holographic gratings" Optica Acta 23, 701-707, 1976 12. Design and Applications Handbook (Photonics 1995) pp.H-372 13. P.T. Rumsby, E.C. Harvey and D.W. Thomas, "Laser Microprojection for Micromechanical Device Fabrication", Proc. SPIE, 2921, 684-692, 1996 14. H. Ohmori, "Electrolytic In-Process Dressing (ELID) Grinding", Int. J. Jpn. Soc. Prec. Eng., 26, 273-278, 1992 15. S. Morita, T. Higuchi, S. Moriyasu, Y. Yamagata, H. Ohmori, N. Ebizuka, "Micro V-groove fabrication for germanium grating by ELID grinding method" (ABTEC'99), Brisbane, Australia, Queensland University of Technology, Society of Grinding Engineers(Japan), 1999 16. N. Ebizuka, S. Morita, H. Ohmori, Y. Yamagata, H. Kobayashi, Y. Hirahara, M. Wakaki and K. Kawaguchi, “Development of immersion gratings for Subaru Telescope by means of ELID grindingmethod”, Advances in Abrasive Technology III, eds. N. Yasunaga, J. Tamaki, K. Suzuki and T. Uematsu, The Soc. of Grinding Engineers Jpn, 149-154, 2000 17. H. Ohmori, N. Ebizuka, S. Morita, Y. Yamagata, "Ultraprecision micro-grinding of germanium immersion grating element for mid-infrared super dispersion spectrograph", CIRP 51st General Assembly, 221-224, 2001 18. N. Ebizuka, S. Morita, Y. Yamagata, H. Ohmori, K. Kawaguchi, M. Wakaki, H. Kobayashi, Y. Hirahara, "Midinfrared high dispersion spectrograph (IHRS) for the Subaru telescope: development of immersion gratings", Diffractive Optics 2001, European Optical Society, 2001
